Quantitative Methods to Guide Epilepsy Surgery After Invasive EEG: A Systematic Review
Abstract number :
3.506
Submission category :
9. Surgery / 9C. All Ages
Year :
2023
Submission ID :
1493
Source :
www.aesnet.org
Presentation date :
12/4/2023 12:00:00 AM
Published date :
Authors :
Presenting Author: Karla Ivankovic, MSc – Universitat Pompeu Fabra
Alessandro Principe, MD, PhD – Universitat Pompeu Fabra; Riccardo Zucca, PhD – Universitat Pompeu Fabra; Rodrigo Rocamora, MD, PhD – Universitat Pompeu Fabra
Rationale:
Many studies reported quantitative EEG analysis methods for improving the precision of surgical targets in drug-resistant epilepsy patients, but none has been implemented in clinical practice yet. The methods are standardly validated on a retrospective cohort, with variable validation parameters. Besides the cohort size, validation parameters are rarely discussed, although they determine the reproducibility and the clinical utility of a method. We reviewed the quality of evidence in reported methods and proposed validation strategies.
Methods:
We searched Scopus for research or conference articles published between 2017 and 2022. Studies validated on invasive EEG in a patient cohort were included. Studies presenting post-surgical outcome prediction without identification of surgical targets were excluded. Extracted parameters included cohort size, surgical outcome, epilepsy types, follow-up and the ground truth reference.
Results:
From 830 studies, 82 met the inclusion criteria. Complete data were available in only five studies, while more than half of the parameters were missing in 14 studies. Cohort size varied from two to 123 patients (median=15) and minimum follow-up from three months to three years (median=one year). Some studies did not include non-seizure-free patients (35%); for those that did, the median ratio of seizure-free to non-seizure-free patients was one. Surgical resection was the ground truth in only 30% of the studies.
Conclusions:
The principal cohort parameters should be clearly stated, as they directly represent the reliability of results. Clinically relevant method must consider a consecutive cohort, with minimal selection criteria. It is important to include non-seizure-free patients and various epilepsy types to ensure the applicability of the method. As the recurrence rate drops after three years, long follow-up ensures result stability. Finally, reports have demonstrated that the ablation of the seizure onset zone is not predictive of good outcome. Thus, we suggest using the resection as ground truth.
Funding:
Ajuts destinats a universitats, centres de recerca i fundacions hospitalàries per contractar personal investigador novell per a l’any 2022 (FI-DGR), Agència de Gestió d’Ajuts Universitaris i de Recerca (AGAUR)
Surgery